This revised application proposes to renew R01 AG030146, Genetic Epidemiology of Cognitive Decline in an Aging Population Sample. Cognitive decline in older age and its most severe manifestation, Alzheimer's disease, are public health problems of enormous magnitude that are projected to become much larger with the continued growth of the oldest population groups. We propose to build on our previous work to further the understanding of the genetic architecture of these important and common phenotypes through an integrated consideration of both epigenomic variation and genomic variation. The proposed work offers substantial potential for increasing our understanding of these phenotypes among African Americans (AAs). An integrated approach to genomic and epigenomic variation may be especially relevant in a study comparing results among AAs and European Americans (EAs) because of the different average exposures to life experiences and environments of these two racial/ethnic groups. The plasticity of the epigenome makes it an excellent place to search for long-lasting traces of past events such as midlife or earlier social and experiential disease risk factors that may be important to understanding racial/ethnic differences in disease risk. We will initially conduct a multi-ethnic meta-analysis of these phenotypes in existing genome wide association scan (GWAS) data from seven cohorts of AAs and EAs representing a large portion of available GWAS data for AAs. We will obtain additional genomic data for a large sample of 5000 AAs and 5000 EAs from these cohorts to characterize the relation of the class of genomic variation in the 0.001-0.05 frequency spectrum to cognitive decline by interrogation of coding and splicing variation throughout the human genome. We will then assemble a comprehensive picture of possible causal genomic variants and of their interconnections using integrated functional dissection of phenomic, transcriptomic (RNA and miRNA), and epigenomic (DNA methylation and H3K9Ac profiles) data available for two of the biracial cohorts. Finally, we will generate DNA methylation profiles and miRNA from 100 AA brains (50 with Alzheimer's disease and 50 controls) from these same two cohorts ) to perform targeted investigations of the role of DNA methylation and miRNA in susceptibility loci of AA subjects and compare these results to those obtained from EA subjects.